Skip to main content
. 2019 Jun 18;30(1):339–352. doi: 10.1093/cercor/bhz091

Table 6.

SEM Tree results for the watershed model

Path Estimate
Before
Age Split
1
Estimate
Between
Age Split
2
Estimate
Between
Age Split
3
Estimate
After
CALM
memory ↔ speed 0.85 8.46 0.97 9.46 0.74
memory → gf 0.83 9.38 0.42 10.04 1.14 10.88 0.94
speed → gf 0.04 6.88 −0.19 11.21 0.17
SLF → memory 0.67 7.21 0.18 11.21 0.76
FMaj → memory 0.59 7.71 0.14 9.29 0.33 11.13 0.74
CG → memory 1 0.64 6.96 0.09 11.04 0.70
ATR → speed 0.96 7.13 0.68 7.96 0.17 11.96 0.65
NKI-RS
memory ↔ speed 0.90 9.82 0.48 14.72 1.11
memory → gf 1.10 8.59 0.59 12.67 1.03
speed → gf 0.53 8.59 −0.12 12.96 0.52
SLF → memory 2.15 8.30 1.47 12.15 1.93
SLF → speed 3.12 8.63 1.83 15.09 2.31

Note. The table shows differences in parameter estimates for paths of interest (as shown in Figs 4 and 5) depending on participants’ age in years. Our analyses allowed for a maximum of 3 age splits (and thus 4 age groups). An absence of a third age split (denoted by"—"in the table) indicates that the SEM tree split only twice, suggesting no further changes in parameter strength after the second split. See Supplementary Figure 7 for a graphical representation of these results.